Car MPG Explorer & Predictor

A simple interactive Shiny application for exploring and predicting automobile fuel economy.

Live application: https://shruthi1232131231.shinyapps.io/mpg_predictor/

Source code: https://github.com/ShruthiRmit/Car-MPG-Explorer-Predictor

The problem

Fuel economy depends on several vehicle characteristics, so a single-variable comparison can be misleading.

The idea

The Car MPG Explorer & Predictor lets a user choose:

• transmission type,
• cylinders,
• vehicle weight, and
• horsepower.

The app then estimates MPG instantly and explains the result in plain language.

Why it is useful

It turns a regression model into something a novice can explore without writing any R code.

The Shiny server fits a multiple linear regression model to R's built-in mtcars data.

cars <- mtcars
cars$am <- factor(cars$am, levels = c(0, 1),
                  labels = c("Automatic", "Manual"))
cars$cyl <- factor(cars$cyl)

fit <- lm(mpg ~ wt + hp + am + cyl, data = cars)

round(summary(fit)$r.squared, 3)

[1] 0.866

The R code above is evaluated when this presentation is rendered.

A higher R-squared indicates that the selected vehicle characteristics explain a substantial share of MPG variation in this small dataset.

Interactive inputs

The user changes sliders and selectors for weight, horsepower, cylinders, and transmission.

Reactive server calculations

Every change triggers a new server-side prediction.

The application displays:

1. Predicted MPG
2. 95% prediction interval
3. Difference from the mtcars average
4. A plot comparing the prediction with the original cars

The app also includes a dedicated How to Use tab so the documentation is available directly on the Shiny site.

Key strengths

• Multiple user-controlled inputs rather than one slider
• A fitted prediction model rather than a simple arithmetic transformation
• Reactive uncertainty estimates
• A visual comparison with the original dataset
• Built-in documentation for novice users
• Clear statement of limitations

Try it

Shiny app: https://shruthi1232131231.shinyapps.io/mpg_predictor/

GitHub code: https://github.com/ShruthiRmit/Car-MPG-Explorer-Predictor

The project demonstrates how statistical modeling, reactivity, and user-centered design can be combined in a compact data product.